INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Myeloperoxidase (MPO), an enzyme found in high abundance in the azurophil granules of neutrophils, catalyzes the hydrogen peroxide-mediated oxidation of the halides chloride, bromide, and iodide and of the pseudohalide thiocyanate to potent microbicidal agents that contribute to the phagocytic antimicrobial armamentarium (18). When chloride is the halide cofactor, the principal oxidant appears to be HOCl (13). The mechanism of microbicidal activity of MPO systems, although extensively studied, remains unclear. Microbicidal activity of the chloride-dependent MPO system occurs concomitantly with a major decline in bacterial DNA synthesis (33), an effect also observed with reagent HOCl (24). In contrast to the MPO microbicidal system, microbicidal effects by gentamicin, a protein synthesis inhibitor, or by an oxidative microbicidal system consisting of acetaldehyde, xanthine oxidase, and iron-EDTA precede the fall in DNA synthesis (33). These findings support a causal role for the suppression of DNA synthesis in the microbicidal effects of the MPO system.

The mechanism of suppression of DNA synthesis by the MPO system is unknown. Diminished DNA synthesis might result from direct DNA oxidation, diminished availability of high-energy nucleotides necessary for DNA assembly (4), or damage to enzymes and structural components responsible for initiation, extension, or termination of chromosomal replication. In Escherichia coli, the initiation of chromosomal DNA synthesis typically originates at a 245-bp region designated the minimal origin of replication (oriC). The replication initiation process, which has been studied extensively (6), involves a large number of cytosolic proteins and appears to require one or more components of the cell envelope (9, 14, 16, 17, 20-23, 26, 27, 36).

The most likely site of the lethal oxidations delivered by the MPO system is the cell envelope. Many envelope components are modified at the time of, or just after, the major decline in bacterial viability (1, 2, 4, 5, 15, 29-32, 34). At the same time, a cytosolic enzyme, like aldolase, is spared within whole bacteria, whereas the cell-free enzyme is highly sensitive to HOCl-mediated inactivation (3). Oxidants may reach the interior of the bacterial cell only after the bacterium is already unable to replicate. These findings favor a view that oxidative events relevant to DNA replication are most likely to occur at the level of the cell envelope rather than at the level of cytosolic DNA oxidation. Of note in this context, MPO-mediated changes in the cell envelope are associated with a decline in the ability of this structure to interact normally with oriC DNA. Further, the degree of decline is proportional to the decline in bacterial DNA synthesis (33).

While cell envelope interactions are a feature of oriC-mediated DNA replication, they have not been described for most other DNA replication origins. We therefore set out to compare episomal DNA synthesis, regulated by a variety of replication origins, to chromosomal DNA replication in E. coli organisms that had been modified by the MPO-mediated antimicrobial system.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Special reagents. Human MPO was prepared and assayed as previously described (30). Glucose oxidase (from Aspergillus niger) (type V-S; Sigma) was used as received, assuming the activity specified by the supplier (1,130 U/ml).

Bacterial strains, plasmids, and growth conditions. E. coli ATCC 11775 was maintained as a frozen stock in 50% (vol/vol) Trypticase soy broth-50% fetal calf serum at 80°C. Plasmids pUC19 (Sigma, St. Louis, Mo.), pACY184 (New England BioLabs), and pSC101 in host E. coli C600 (ATCC 37032) were obtained from commercial sources. Plasmid pSP102 (28) was a gift from D. Chattoraj. Plasmid pCM700, obtained from R. Allen, Department of Biochemistry, Stanford University, was derived from the minichromosome pCM959 (8, 25) by excision of an 837-bp PvuII fragment followed by ligation insertion of a chloramphenicol acetyltransferase gene conferring chloramphenicol resistance. Where indicated, plasmids were isolated from host strains and transformed into E. coli ATCC 11775. Transformed strains were stored at 80°C in glycerolized Luria Bertani broth as previously described (35).

Microbicidal assay. E. coli organisms containing the plasmids of interest were grown overnight at 37°C with vigorous agitation in 100 ml of Trypticase soy broth supplemented with the appropriate antibiotic (ampicillin, 100 µg/ml; chloramphenicol, 50 µg/ml; or tetracycline, 10 µg/ml). Organisms were centrifuged for 15 min at 3,000 × g and 4°C and were washed once with 10 ml of 0.1 M Na₂SO₄ supplemented with 0.1% gelatin. The suspension was adjusted to a turbidity characteristic of organisms at 10 times the final desired concentration and was maintained on ice.

DNA synthesis. DNA synthesis was estimated as assimilation of tritiated thymidine. A 2.5-ml portion of the microbicidal reaction mixture was mixed with an equal volume of double-strength trypticase soy broth that had been prewarmed to 37°C and supplemented with 2 mCi of [³H]thymidine/ml. The suspension was tumbled for 45 min in an incubator at 37°C, after which further thymidine incorporation was inhibited by the addition of ciprofloxacin (50 µg/ml). Tubes containing the reaction mixture were mixed and stored on ice.

Data analysis. Data were expressed as percentages of a baseline value specific to each experiment. The baseline values were the averages of the respective viability, plasmid radioactivity, and total radioactivity values obtained from organisms harvested 1 min prior to, immediately after, and 2 min after addition of MPO. This approach was adopted to minimize the effects that a single aberrant value might have on the percentages determined for an entire experiment.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

It was previously observed (33) that the decline of DNA synthesis in MPO-treated E. coli nearly matched declines in viability, measured by the ability to form colonies in nutrient agar. Figure 1 shows similar data obtained from all E. coli preparations described in this report. Immediately after a period of exposure to the MPO microbicidal system, cells were diluted in growth medium supplemented with tritiated thymidine and cultured at 37°C for 45 min. DNA synthesis was assessed as incorporation of radioactivity into alcohol-insoluble material. Analysis of the relationship between MPO effects on microbial viability and those on total DNA synthesis indicated a statistically significant deviation from linearity (P = 0.01 for the quadratic term of the regression analysis). However, the magnitude of this term was small. Of the total variation, 81% could be ascribed to the linear component, 18% to random error, and <1% to the quadratic component.

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Relationships among chromosomal DNA synthesis and bacterial viability in MPO-treated E. coli organisms. Total DNA synthesis (100% tritiated thymidine incorporation, 212 ± 78 [mean ± standard deviation] nmol per 109 E. coli per 45 min [mditn = 28]) was compared to residual bacterial viability (100%, 1.1 ± 0.3 × 109 cells per ml) for all experiments described in Fig. 2 and 3. Data (n = 280) were grouped as described in Materials and Methods. Error bars represent standard errors of the means. Eight data points ranged outside the graph axes and, while included in the groupings, are not shown individually.

In addition to evaluating total thymidine incorporation (total DNA synthesis), plasmid DNA was isolated to assess plasmid specific uptake. For moderate-copy-number plasmids in this study (<20 copies per cell), i.e., all but pUC19, plasmid thymidine incorporation accounted for 1.4% ± 0.7% (mean ± standard deviation, n = 23) of total uptake. For pUC19 (200 to 500 copies per cell), plasmid thymidine incorporation was 7.9% ± 3.0% (n = 5) of the total. Since the bulk of total thymidine incorporation was not plasmid associated, total thymidine uptake was used to estimate chromosomal DNA synthesis.

Figure 2 describes the relationships among plasmid DNA synthesis and chromosomal DNA synthesis for two episomes, pCM700 and pUC19, after various periods of incubation with the MPO system. Plasmid pCM700 is derived from a minichromosome that replicates from the chromosomal origin, oriC. The decline in pCM700 DNA synthesis, described in Fig. 2a, closely matched the decline in total cell DNA synthesis. The slope of the regression line for 50 individual data points was 0.90 ± 0.05 (standard error), with an intercept of 0.08 ± 0.04 and on r² of 0.89 ± 0.13. Figure 2b describes similarly obtained data for plasmid pUC19, which replicates from the ColE1-like pMB origin of replication. Replication of this plasmid was relatively resistant to inactivation by the MPO system and appeared to be normal in E. coli cells that had lost as much as 50% of their initial capacity to synthesize DNA. Regression analysis indicated that the deviation from a linear relationship between plasmid and chromosomal DNA synthesis was highly significant (P < 0.0001) (Table 1).

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Relationships among plasmid and chromosomal DNA synthesis in MPO-treated E. coli (pCM700 or pUC19). Plasmid-containing E. coli cells were incubated with the MPO microbicidal system, and at intervals samples were removed for determination of plasmid and total DNA syntheses as described in Materials and Methods. (A) Plasmid pCM700 containing the chromosomal origin of replication oriC. (B) plasmid pUC19 containing the pMB1 origin of replication. Individual data points were grouped as described in Materials and Methods. Error bars represent standard errors of the means.

Three other plasmids were evaluated for their ability to maintain DNA synthesis within E. coli after exposure to MPO-derived oxidants (Fig. 3). These were pSP102, with an origin derived from the P1 phage; pSC101, with an origin absolutely dependent on several properties of the dnaA protein; and pACYC184, with the p15a origin of replication, similar to ColE1, but with a lower copy number than pUC19. pSP102 exhibited a pattern of inactivation that most nearly resembled that of pCM700 and thus matched the inactivation rate of chromosomal DNA synthesis. The slope of the regression line for 33 data points (dropping an extreme outlier) was 0.95 ± 0.04, with an intercept of 0.11 ± 0.05, and an r² of 0.92 ± 0.17 (inclusion of the outlier resulted in an r² of 0.87). The remaining plasmids exhibited a pattern of inactivation that was most like that of pUC19 (P  0.0001 for deviation from linearity) (Table 1), indicating relative resistance to oxidative inactivation by the MPO system.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   Relationships among plasmid and chromosomal DNA synthesis in MPO-treated E. coli (multiple plasmids). Conditions were as described in the legend for Fig. 2. , pCM700 (n = 50); , pSP102 (n = 34); , pUC19 (n = 48); , pACYC184 (n = 36); , pSC101 (n = 39). Only grouped data are shown and error bars are omitted.

In preliminary studies to assess the direct attack of plasmid DNA by MPO-derived oxidants, plasmid preparations from MPO-treated E. coli cells were analyzed by agarose gel electrophoresis without prior restriction enzyme digestion. The host E. coli strain was endonuclease negative (DH5). Reasoning that oxidative nicking might produce a conversion from supercoiled DNA to relaxed circles, or even linearized plasmids, which have distinct migration characteristics in agarose, the relative abundance of each form was examined by direct inspection of ethidium bromide-stained gels. While there were readily detectable background levels of relaxed circles and faint indications of linearized plasmid, exposure to lethal amounts of MPO-derived oxidants failed to alter the distribution of plasmid forms for either pCM700- or pUC19-containing E. coli (data not shown).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

As previously observed (33), the MPO-dependent antimicrobial system rapidly and extensively impairs the capacity of E. coli to synthesize new DNA. The current data exhibited, in contrast to earlier observations, a statistically significant (P = 0.01) nonlinear component to the relationship between loss of viability and loss of DNA synthesis. However, the magnitude of the nonlinear coefficient was small and is, in our view, of dubious biological significance. We continue to believe that there is an essentially linear relationship between these two MPO-mediated effects.

This study used incorporation of thymidine into macromolecules (total DNA synthesis) an estimate of chromosomal DNA synthesis. How accurate is this estimate? Prior work (10) indicated that for E. coli containing the ColE1 plasmid, more than 98% of total DNA synthesis can be attributed to the chromosome. These observations were affirmed in this study for plasmids with copy numbers below 20 per cell. Even for pUC19, with a copy number of 500 to 700 plasmids per cell (35), 93% of total DNA synthesis could be attributed to chromosomal replication. Thus, equating total DNA synthesis to chromosomal DNA synthesis introduces only modest errors, even for high-copy-number plasmids.

The general comparison of plasmid to chromosomal DNA replication following MPO exposure indicated two patterns of response. Plasmids pCM700 (oriC replicon) and pSP102 (P1-oriR replicon) (group I) lost the ability to synthesize new DNA at rates that were similar to those for total DNA. In contrast, plasmids pUC19, pACYC184, and pSC101 (group II) were protected from the early effects of MPO-derived oxidants. Loss of DNA synthetic ability for group II plasmids declined rapidly only after 50 to 70% of chromosome-replicating activity had been lost, and the deviation from a linear relationship was highly significant (Table 1, P  0.0001).

Several features of the DNA replication requirements for the plasmids employed in this study are summarized in Table 1. The relative resistance of pUC19 and pACYC184 to MPO inactivation suggests that RNA polymerase, DNA polymerase I, and the dnaB, dnaC, dnaG, dnaE, and dnaT proteins are not inactivated early. Further, the resistance of pSC101, which is sensitive to a broad array of mutations in the dnaA gene (19) suggests that the dnaA protein is also not a critical target for MPO-mediated oxidations at the early stages of killing.

While depletion of high-energy nucleotides (4) might account for a decline in DNA synthesis, it is difficult to reconcile the divergent effects on the two plasmid groups based on such a global impairment to DNA synthesis. Although similar concerns might apply to considerations about direct oxidation of DNA polymers by MPO-derived oxidants, it is noteworthy that the topologies of group I and group II plasmids appear to differ. Both oriC and P1 plasmids are sequestered at the membrane for a significant period of time during each cell cycle (12, 26, 37) through the intermediacy of the SeqA protein. It is conceivable that apposition to the outer envelope of the cell renders these membrane-associated regions of DNA more susceptible to externally generated oxidants. Studies of plasmids recovered from MPO-oxidized E. coli containing either the cell envelope-interactive plasmid pCM700 or the noninteractive plasmid pUC19 failed to identify major conversions of supercoiled plasmid DNA to relaxed or linear forms. However, more sophisticated studies are in order before this issue can be addressed adequately.

In summary, an analysis of the effects of MPO-derived oxidants on synthesis of plasmid DNA has indicated that oriC- and P1-based plasmids are more susceptible than ColE1-type (pMB1 and p15a) and pSC101 plasmids to MPO inactivation. We envision the following scenario for the differential MPO effects on chromosome, minichromosome, and P1 origin episomes on the one hand and those on more resistant episomes on the other. MPO-derived oxidants attack all susceptible targets in their path, beginning at the outside of the bacterial cell where they are generated and working inwards. Soon, cell envelope structures essential for efficient replication of chromosome, minichromosome, and P1 DNA are disrupted. It is proposed that these components are not essential for replication of the resistant episomes. For these episomes DNA synthesis persists until MPO-mediated oxidation becomes so extensive as to produce global impairment of DNA synthesis, perhaps by depletion of high-energy nucleotide substrates (4), resulting in a delayed but precipitous decline in DNA synthesis for the resistant episomes as well. Support for these conjectures would depend on identification of high levels of MPO susceptibility in situ among replication factors specific to the oriC and P1 replicons, a focus of continuing investigation.